Source: Mike Brennan, Partner, CSC
For many financial sector firms, data management and integration problems may only be resolved through the adoption of an enterprise-wide service oriented architecture says Mike Brennan, a partner with CSC's consulting practice.
When it comes to the applications that support the Financial Services Industry, “buy versus build” has been the operating environment for some time now. While a few diehards remain who actually believe they can build better and less expensive systems than the leading vendors, overall the prevailing wisdom is to obtain them from the folks who create and service them for a living.
As a result, the best description of application maps for many organizations is a loose federation of vendor packages with some remaining proprietary components.
Which prompts the question: how well are they integrated?
Therein lies the problem. Industry participants who were at first excited about the upgrades vendor packages provided are now wrestling with the aftermath of implementation.
The common industry participant structure of enterprise, customer and product silos, combined with best of breed applications, has created serious integration and data management issues.
Considering the wide reaching impact, these are clearly the biggest challenges the industry is currently facing. Nowhere is this more apparent than in two frequently mentioned pain points: (a) complying with new and demanding regulatory requirements; and (b) obtaining a single view of a customer. Both require integration and data management capabilities most industry participants don’t have and are rushing to provide.
Interim strategies revolve around the extensive use of excel spreadsheets and data warehouses. Some participants have reached the point where data warehouses are receiving data exclusively from other data warehouses. It’s frightening that this data finds its way to places like customer statements and general ledgers, with C level executives attesting to the accuracy. This is not a formula for either regulatory compliance or customer service.
Unfortunately, there is no quick fix — and for this problem, “buy versus build” does not apply. What does apply is achieving integration and data management capabilities as part of a broader IT transformation strategy focused on the development of a Service-Oriented Architecture (SOA).
Moving Towards Service-Oriented Architecture
Incorporating these challenges into a transformation initiative focused on addressing enterprise wide cost, growth and risk concerns is a sensible approach with a much greater chance of success than standalone initiatives. That said, it is important to understand that this is a considerable undertaking in terms of expense, risk and organizational change. No doubt, some organizations will look for other alternatives ranging from living with what they have to applying an outsourcing strategy.
For those choosing to transform, these key characteristics of an SOA-based IT Transformation should be considered:
1) Understanding that Service Level Architecture = Web Services (i.e. XML messaging and web protocols).
2) Creating a business process-oriented transformation vision with business processes configured in one central function — a workflow engine — rather than hard coded in multiple applications.
3) Performing analysis that addresses both the business processes and core IT components: technical, application and data.
4) Assessing the degree of change between present state and future state.
5) Identifying a high-level approach for filling the gaps.
6) Identifying key business and functional requirements.
7) Identifying reusable assets.
8) Developing a transformation architecture approach defined as the set of architectures associated with each release. This includes the architecture plus a mapping of conceptual components from the present state to the future state and an asset retirement strategy identifying assets to be retired at each release.
9) Enhancing release plans to include transformation and transformation issues and techniques.
10) Developing an operational support plan for release transition.
The end result will be an architecture that is low cost, scalable, and application agnostic — with the ability to manage data no matter where it resides.
While an SOA-based IT Transformation is a relatively new concept, the inflexible demands of the industry require it to be considered. The inability to meet those demands puts industry participants in difficult positions. No one wants to be in the position of explaining non-compliance to regulatory authorities or defending poor service to loyal customers. Conversely, no one wants to listen to those explanations.